Geometric 4D Stitching for Grounded 4D Generation
2026-05-11 • Computer Vision and Pattern Recognition
Computer Vision and Pattern RecognitionArtificial IntelligenceMachine Learning
AI summaryⓘ
The authors address problems in creating 4D scenes, where current methods struggle with inconsistent shapes and require slow optimization. They introduce Geometric 4D Stitching, a faster way to find and fix missing parts in 3D scenes over time, improving shape accuracy. Their method works quickly on a single GPU and allows easy updates and edits to 4D scene meshes.
4D generationgeometric consistencyradiance-based representationgenerative models3D meshscene reconstructionNVIDIA RTX 5090optimizationview-dependent effectsscene editing
Authors
Sunwoo Park, Taesung Kwon, Jong Chul Ye
Abstract
Recent 4D generation methods complete scene-level missing information using generative models and reconstruct the scene into radiance-based representations. However, these pipelines often present geometric inconsistencies in the generated content, and the radiance-based reconstruction requires expensive optimization. Furthermore, radiance-based representations often absorb these geometric inconsistencies into their view-dependent nature, failing to enforce the grounded geometric consistency. To address these issues, we propose Geometric 4D Stitching, an efficient framework that explicitly identifies missing geometric regions and complements them with geometrically grounded 4D stitches. As a result, our method constructs 4D scene representations in under 10 minutes on a single NVIDIA RTX 5090 GPU per one-step scene expansion, while improving geometric consistency. Moreover, we demonstrate that our explicit 4D stitching supports interative expansion of 4D mesh as well as 4D scene editing.